一种基于查询重写数据服务组合方法.pdfVIP

一种基于查询重写的数据服务组合方法 * 周林华 ，陈华钧，毛郁欣 (浙江大学 计算机学院，浙江 杭州 310027) 摘 要：当前产品研发产生的大量数据需要进行聚合才能满足研究人员的研究目标，数据服务组合已 成为解决数据聚合的关键技术。介绍了一种基于本体的方法来发布与组织数据密集型的 Web 服务， 使用语义 Web 规范描述 Web 服务的内容及能力，提出动态演变的 Web 服务接口技术和一个基于查询 重写计划的服务组合算法。该方法的优点是数据服务能够动态适应用户的查询请求，从而提高用户查 询请求的满足度。模拟实验表明该方法能够较好地适应大规模的数据服务组合情况，通过与固定 WSDL 接口的数据服务比较，说明该方法更适用于用户查询请求交互度较高的场合。 关键词：数据服务；服务组合；查询重写；本体 中图分类号: TP393 文献标识码: A A data service composition approach based-on query rewriting ZHOU Lin-hua*, CHEN Hua-jun, MAO Yu-xing ( School of Computer Science, Zhejiang University, Hangzhou 310027, China) Modern product research is generating daunting amount of data that is commonly required to be combined Abstract ： together to service daily research endeavors. Data Service composition have become a key technology for enabling this kinds of data combination. This paper introduces an ontology-based approach to publishing and composing data intensive web service. The approach distinguishes itself by its capability of dynamically evolving service interface and a query rewriting-based planner for service composition, enabled by richly describing service capability using the semantic web languages. The advantage of the proposed approach is that the data service can adapt itself with client request dynamically, and improving the user’s satisfaction of query requests. The evaluation reveals the approach scale well as service number goes large, and is applicable to the interaction situation comparing with the Data Service of having fixed interfaces. Key words ：data service; service composition; query rewriting